Meta changes the way its AI chatbot responds to kids after senator launches probe into its conversations with teens

Meta has announced it is modifying its AI chatbot's responses to ensure a safer and more age-appropriate experience for teenagers. The company's decision comes in light of a recent report that revealed internal documents suggesting the chatbot could engage in romantic dialogues with minors. In a statement provided to Business Insider, a Meta spokesperson indicated that these adjustments are temporary measures aimed at enhancing safety while the company works on more robust long-term strategies. Stephanie Otway emphasized that the modifications involve adding protective measures, such as training the AI to avoid sensitive topics and directing users to expert resources. The list of restricted subjects includes not only romantic interactions but also critical issues like self-harm, suicide, and eating disorders. According to Otway, the AI characters accessible to teenagers will focus on educational and creative expression rather than sensitive conversations. The urgency of these changes was highlighted by Senator Josh Hawley, who announced an investigation into Meta’s practices after the unsettling revelations. In a letter to CEO Mark Zuckerberg, Hawley criticized the company for allowing chatbots to engage in potentially harmful interactions with children, stating that action was only taken after the company faced public scrutiny. Further complicating the matter, the nonprofit group Common Sense Media issued a warning against the use of Meta's AI chatbot for anyone under 18, citing findings that these AI tools often mislead teens and may promote harmful behaviors. This incident adds to a growing list of concerns regarding Meta's handling of children's safety, following Zuckerberg's previous testimony alongside other tech leaders about the risks posed by social media platforms.